Saved in:
Bibliographic Details
Main Authors: Kim, Taewoon, François-Lavet, Vincent, Cochez, Michael
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.22142
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911704293572608
author Kim, Taewoon
François-Lavet, Vincent
Cochez, Michael
author_facet Kim, Taewoon
François-Lavet, Vincent
Cochez, Michael
contents Reinforcement learning under partial observability requires deciding what information to retain, yet most memory-based approaches do not explicitly model short-term-to-long-term transfer of symbolic observations. We study this transfer process in a temporal knowledge-graph memory setting and cast it as a neuro-symbolic value-based decision problem: for each observed triple, the agent chooses whether to keep or drop it before long-term insertion. To handle variable-sized short-term buffers, we use a per-item Q-learning design with shared parameters and a practical temporal-difference update over matched items across consecutive steps. On the RoomKG benchmark at long-term memory capacity 128, learned transfer decisions outperform symbolic and neural baselines, including symbolic baselines with temporal annotations and history-based LSTM/Transformer baselines. Across transfer-policy ablations, a lightweight local short-term-only variant performs best, and step-level behavior shows that the policy keeps navigation- and query-relevant facts while discarding lower-value candidate facts, supporting explicit and interpretable memory decisions under memory constraints.
format Preprint
id arxiv_https___arxiv_org_abs_2605_22142
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Short-Term-to-Long-Term Memory Transfer for Knowledge Graphs under Partial Observability
Kim, Taewoon
François-Lavet, Vincent
Cochez, Michael
Machine Learning
Artificial Intelligence
Reinforcement learning under partial observability requires deciding what information to retain, yet most memory-based approaches do not explicitly model short-term-to-long-term transfer of symbolic observations. We study this transfer process in a temporal knowledge-graph memory setting and cast it as a neuro-symbolic value-based decision problem: for each observed triple, the agent chooses whether to keep or drop it before long-term insertion. To handle variable-sized short-term buffers, we use a per-item Q-learning design with shared parameters and a practical temporal-difference update over matched items across consecutive steps. On the RoomKG benchmark at long-term memory capacity 128, learned transfer decisions outperform symbolic and neural baselines, including symbolic baselines with temporal annotations and history-based LSTM/Transformer baselines. Across transfer-policy ablations, a lightweight local short-term-only variant performs best, and step-level behavior shows that the policy keeps navigation- and query-relevant facts while discarding lower-value candidate facts, supporting explicit and interpretable memory decisions under memory constraints.
title Short-Term-to-Long-Term Memory Transfer for Knowledge Graphs under Partial Observability
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2605.22142